Hiring a Power BI consultant: what should an SME look for?

What does a Power BI consultant actually do?

The title covers a broad field. In practice, a Power BI consultant working for an SME usually does four things:

• Sharpening the question. Which decision do you want to make with the dashboard, and which figures do you need for it? This is also where you pin down what "revenue" or "margin" actually means.
• Getting to the data. Connecting Power BI to your accounting, CRM, ERP or HR software, and making sure that data refreshes automatically.
• Building the data model. Setting up the tables so that customers, products, periods and locations link up properly and the calculations are correct.
• Building and handing over the dashboard. An overview with filters and detail pages, tested with the people who will use it, plus explanation and documentation.

The first and third points are the ones most often underestimated. A good-looking dashboard on a messy model gives good-looking but wrong numbers.

When does an SME need a Power BI consultant?

Not always. If you have one system with ready-made reporting that does what you want, or someone on your team enjoys Power BI and is given time for it, you can get quite far on your own. Our article Excel or Power BI helps you judge whether switching is needed at all.

A consultant becomes worthwhile when one or more of these sound familiar:

1. Your data comes from several systems. Revenue from accounting, opportunities from the CRM, hours from planning. Linking them is the hardest part of the job.
2. You need a management or team dashboard that people actually steer by. Then the numbers have to be right and everyone has to use the same definition.
3. You already have a report, but nobody trusts it. Figures differ from the accounts, refreshes fail or the report is so slow nobody opens it.
4. You want results quickly and have no time to learn the craft. Power BI is easy to start with, but the data model and the formulas (DAX) take experience.

Freelancer, large agency or small specialist: what suits an SME?

There are roughly three options, each with its own pros and cons:

• Freelance Power BI consultant. Often flexible and directly reachable. Watch continuity: what happens when they are fully booked or ill, and is the work documented well enough for someone else to take over?
• Large consultancy. Plenty of capacity and broad expertise, but often geared to larger organisations with their own IT department. For an SME question, the project can become heavier than needed.
• Small specialist. Short lines, often with experience of the connections and the data foundation beneath the dashboard. Do ask about the way of working and who actually does the work.

More important than the form is whether someone understands how your systems fit together. A consultant who already knows your accounting or CRM software has less to figure out.

Which questions should you ask in the first conversation?

These questions quickly show whether someone is a good fit:

1. "Which question will we answer first?" A good consultant wants to know which decision you make with the dashboard before talking about visuals.
2. "How will you get to my data?" Is there an existing connector for your software, does one need to be built, or does it run on exports? That determines how reliable and low-maintenance it will be.
3. "How will we record the definitions?" What is revenue: invoiced, paid or booked? If this is not written down, you will argue about it later.
4. "What do I get at handover?" Documentation of the model and calculations, access and admin rights, and an explanation for your team.
5. "What about licences?" Who needs which Power BI licence to view or share the dashboard? To prepare, read our article on Power BI licensing.
6. "What is the scope and price per phase?" A fixed agreement per phase gives more control than an open-ended hourly arrangement.

An example: how a first project runs

A fictional but recognisable example. A wholesaler with around twenty-five employees wants to know which customers and product groups bring in the most margin. Revenue sits in the ERP, purchase prices partly in Excel, and account managers track their customers in a CRM.

1. Question and definitions. In a first session it is agreed that the focus is gross margin per customer and per product group, per month, based on invoiced orders. It is also agreed which Excel list is leading for purchase prices.
2. Getting to the data. The ERP and CRM are connected so they refresh automatically. For purchase prices, it is agreed they will from now on live in one fixed file with a fixed layout.
3. Model and checks. The data model is built and the results are compared with the accounts. Differences are investigated before anyone sees the dashboard.
4. Dashboard and handover. One overview page with margin per customer and product group, plus a detail page per customer. Management and two account managers test it, there is a short walkthrough and the documentation is handed over.

Afterwards there is no big BI platform, but there is one dashboard that is correct and used. Only then is it time for the next question. The figures that often come up next are covered in the 10 KPIs every SME dashboard needs.

How do you avoid becoming dependent on your consultant?

This may be the most important point. A dashboard that only its builder understands is a risk. So agree upfront that:

• The report and connections live in your environment. That means your own Microsoft tenant and your own accounts, not the consultant's.
• You have admin rights. Someone inside your company must be able to open, change and share the report.
• The model and calculations are documented. So a colleague or another party can take it over later.
• Maintenance is a separate choice. It must be possible to carry on yourself after handover. Ongoing maintenance can be useful, but it should not be a condition.

What mistakes are common when hiring a Power BI consultant?

• Starting with a wish list of twenty charts. Start with one question. A small dashboard that is correct gets used sooner than a big one that is half finished.
• Skipping the data model. If the focus is only on looks, you will pay later to fix the foundations.
• Not freeing up a colleague. The consultant knows Power BI, your colleague knows the process. Without someone reviewing and testing, you miss the edge cases.
• Looking at licences only at the end. Then it turns out colleagues cannot open the report. Discuss this at the start.

How does Celsius BI approach this?

We build Power BI dashboards for SMEs, on the data you already have. We start from the question, agree the KPI definitions upfront and check the results with your team before the dashboard goes live. You get documentation of the model and calculations, and the dashboard and its connections are handed over including access and admin rights. After the first conversation you receive a proposal with a fixed price per phase.

Read more about the approach under dashboarding and our approach. The systems we can already connect to Power BI are listed under integrations. If your data first needs to be brought together, that falls under data engineering.